Feasibility of Emotions as Features for Suicide Ideation Detection in Social Media

Computing and Communications

Associated organisational unit

Data Science

Electronic data

HealTAC2023_Paper
Accepted author manuscript, 136 KB, PDF document
Available under license: CC BY: Creative Commons Attribution 4.0 International License

Keywords

Natural Language Processing, Suicide Ideation Detection, Social Media Analytics, Emotion Detection, Large Language Models, Machine Learning

View graph of relations

Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review

Published

Standard

Feasibility of Emotions as Features for Suicide Ideation Detection in Social Media. / Arreerard, Ratchakrit ; Piao, Scott.
2023. Paper presented at HEALTHCARE TEXT ANALYTICS CONFERENCE 2023, Manchester, United Kingdom.

Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review

Harvard

Arreerard, R & Piao, S 2023, 'Feasibility of Emotions as Features for Suicide Ideation Detection in Social Media', Paper presented at HEALTHCARE TEXT ANALYTICS CONFERENCE 2023, Manchester, United Kingdom, 15/06/23 - 16/06/23.

Bibtex

@conference{0da4cb292b4c4e95966ab7952c484a2f,

title = "Feasibility of Emotions as Features for Suicide Ideation Detection in Social Media",

abstract = "Suicide-related social media message detection is an important issue. Such messages can reveal a warning sign of suicidal behaviour. This paper examines the efficacy of using emotions as sole features to detect suicide-related messages. We investigated two methods which use a single emotion and a set of seven emotions as features respectively. For emotion classification, we used a classifier based on BERT named {"}Emotion English DistilRoBERTa-base{"}. For detecting suicide-related messages, we tested Naive Bayes and Support Vector Machine. As our training/test data for suicide message detection, we used a publicly available dataset collected from Reddit in which each post is labelled as {"}suicide{"} or {"}non-suicide{"}. Ourmethod obtained accuracies of 76.2% and 76.8% for detecting suicide-related messages with Naive Bayes and Support Vector Machine respectively. Our experiment also shows that three emotion categories, {"}anger{"}, {"}fear{"} and {"}sadness{"}, have a strongest correlation with suicide-related messages.",

keywords = "Natural Language Processing, Suicide Ideation Detection, Social Media Analytics, Emotion Detection, Large Language Models, Machine Learning",

author = "Ratchakrit Arreerard and Scott Piao",

year = "2023",

month = jun,

day = "16",

language = "English",

note = "HEALTHCARE TEXT ANALYTICS CONFERENCE 2023 : PhD Forum, HealTac 2023 ; Conference date: 15-06-2023 Through 16-06-2023",

url = "http://healtex.org/healtac-2023/",

}

RIS

TY - CONF

T1 - Feasibility of Emotions as Features for Suicide Ideation Detection in Social Media

AU - Arreerard, Ratchakrit

AU - Piao, Scott

PY - 2023/6/16

Y1 - 2023/6/16

N2 - Suicide-related social media message detection is an important issue. Such messages can reveal a warning sign of suicidal behaviour. This paper examines the efficacy of using emotions as sole features to detect suicide-related messages. We investigated two methods which use a single emotion and a set of seven emotions as features respectively. For emotion classification, we used a classifier based on BERT named "Emotion English DistilRoBERTa-base". For detecting suicide-related messages, we tested Naive Bayes and Support Vector Machine. As our training/test data for suicide message detection, we used a publicly available dataset collected from Reddit in which each post is labelled as "suicide" or "non-suicide". Ourmethod obtained accuracies of 76.2% and 76.8% for detecting suicide-related messages with Naive Bayes and Support Vector Machine respectively. Our experiment also shows that three emotion categories, "anger", "fear" and "sadness", have a strongest correlation with suicide-related messages.

AB - Suicide-related social media message detection is an important issue. Such messages can reveal a warning sign of suicidal behaviour. This paper examines the efficacy of using emotions as sole features to detect suicide-related messages. We investigated two methods which use a single emotion and a set of seven emotions as features respectively. For emotion classification, we used a classifier based on BERT named "Emotion English DistilRoBERTa-base". For detecting suicide-related messages, we tested Naive Bayes and Support Vector Machine. As our training/test data for suicide message detection, we used a publicly available dataset collected from Reddit in which each post is labelled as "suicide" or "non-suicide". Ourmethod obtained accuracies of 76.2% and 76.8% for detecting suicide-related messages with Naive Bayes and Support Vector Machine respectively. Our experiment also shows that three emotion categories, "anger", "fear" and "sadness", have a strongest correlation with suicide-related messages.

KW - Natural Language Processing

KW - Suicide Ideation Detection

KW - Social Media Analytics

KW - Emotion Detection

KW - Large Language Models

KW - Machine Learning

M3 - Conference paper

T2 - HEALTHCARE TEXT ANALYTICS CONFERENCE 2023

Y2 - 15 June 2023 through 16 June 2023

ER -

Research

Associated organisational unit

Electronic data

Keywords